Evaluating the efficacy of planktonic foraminifer calcite d 18 O data for sea surface temperature reconstruction for the Late Miocene Évaluation de l’efficacité du d 18 O de la calcite des foraminifères planctoniques pour la reconstitution des températures de la surface des mers au Miocène supérieur Mark Williams a, *, Alan M. Haywood a , Steve P. Taylor b , Paul J. Valdes c , Bruce W. Sellwood d , Claus-Dieter Hillenbrand a a British Antarctic Survey, Geological Sciences Division, High Cross, Madingley Road, Cambridge CB3 0ET, United Kingdom b Department of Meteorology, University of Reading, Earley Gate, PO Box 243, Reading RG6 6BB, United Kingdom c School of Geographical Sciences, University of Bristol, University Road, Bristol BS8 1SS, United Kingdom d School of Human and Environmental Sciences, University PO Box 227, Whiteknights, Reading RG6 6AB, United Kingdom Received 2 September 2004; accepted 22 December 2004 Available online 21 October 2005 Abstract This study examines the efficacy of published d 18 O data from the calcite of Late Miocene surface dwelling planktonic foraminifer shells, for sea surface temperature estimates for the pre-Quaternary. The data are from 33 Late Miocene (Messinian) marine sites from a modern latitudinal gradient of 64°N to 48°S. They give estimates of SSTs in the tropics/subtropics (to 30°N and S) that are mostly cooler than present. Possible causes of this temperature discrepancy are ecological factors (e.g. calcification of shells at levels below the ocean mixed layer), taphonomic effects (e.g. diagenesis or dissolution), inaccurate estimation of Late Miocene seawater oxygen isotope composition, or a real Late Miocene cool climate. The scale of apparent cooling in the tropics suggests that the SST signal of the foraminifer calcite has been reset, at least in part, by early diagenetic calcite with higher d 18 O, formed in the foraminifer shells in cool sea bottom pore waters, probably coupled with the effects of calcite formed below the mixed layer during the life of the foraminifera. This hypothesis is supported by the markedly cooler SST estimates from low latitudes—in some cases more than 9 °C cooler than present—where the gradients of temperature and the d 18 O composition of seawater between sea surface and sea bottom are most marked, and where ocean surface stratification is high. At higher latitudes, particularly N and S of 30°, the temperature signal is still cooler, though maximum temperature estimates overlap with modern SSTs N and S of 40°. Comparison of SST estimates for the Late Miocene from alkenone unsaturation analysis from the eastern tropical Atlantic at Ocean Drilling Program (ODP) Site 958—which suggest a warmer sea surface by 2–4 °C, with estimates from oxygen isotopes at Deep Sea Drilling Project (DSDP) Site 366 and ODP Site 959, indicating cooler than present SSTs, also suggest a significant impact on the d 18 O signal. Nevertheless, much of the original SST variation is clearly preserved in the primary calcite formed in the mixed layer, and records secular and temporal oceanographic changes at the sea surface, such as movement of the Antarctic Polar Front in the Southern Ocean. Cooler SSTs in the tropics and sub-tropics are also consistent with the Late Miocene latitude reduction in the coral reef belt and with interrupted reef growth on the Queensland Plateau of eastern Australia, though it is not possible to quantify absolute SSTs with the existing oxygen isotope data. Reconstruction of an accurate global SST dataset for Neogene time-slices from the existing published DSDP/ODP isotope data, for use in general circulation models, may require a detailed re-assessment of taphonomy at many sites. © 2005 Elsevier SAS. All rights reserved. * Corresponding author. School of Earth and Environmental Sciences, University of Portsmouth, Burnaby Building, Burnaby Road, Portsmouth PO1 3QL, United Kingdom. E-mail address: mark.williams@port.ac.uk (M. Williams). Geobios 38 (2005) 843–863 http://france.elsevier.com/direct/GEOBIO/ 0016-6995/$ - see front matter © 2005 Elsevier SAS. All rights reserved. doi:10.1016/j.geobios.2004.12.001